Python 深拷贝和浅拷贝详解

1、Python 深拷贝和浅拷贝概念理解

• 浅拷贝，指的是重新分配一块内存，创建一个新的对象，但里面的元素是原对象中各个子对象的引用。

• 深拷贝，是指重新分配一块内存，创建一个新的对象，并且将原对象中的元素，以递归的方式，通过创建新的子对象拷贝到新对象中。因此，新对象和原对象没有任何关联。

2、浅拷贝

• 使用数据类型本身的构造器
• 对于可变的序列，还可以通过切片操作符 : 来完成浅拷贝
• Python 还提供了对应的函数 copy.copy() 函数，适用于任何数据类型

2.1 使用数据类型本身的构造器

list1 = [1, 2, 3]
list2 = list(list1)
print(list2)
print("list1==list2 ?",list1==list2)
print("list1 is list2 ?",list1 is list2)

set1= set([1, 2, 3])
set2 = set(set1)
print(set2)
print("set1==set2 ?",set1==set2)
print("set1 is set2 ?",set1 is set2)

dict1 = {1:[1,'w'], 2:0, 3:98}
dict2 = dict(dict1)
print(dict2)
print("dict1 == dict2 ?",dict1 == dict2)
print("dict1 is dict2 ?",dict1 is dict2)

[1, 2, 3]
list1==list2 ? True
list1 is list2 ? False

{1, 2, 3}
set1==set2 ? True
set1 is set2 ? False

{1: [1, 'w'], 2: 0, 3: 98}
dict1 == dict2 ? True
dict1 is dict2 ? False

分析： 浅拷贝，为新变量重新分配一块内存，和原来变量的内存不一样，所以有

list1 is list2 ? False
set1 is set2 ? False
dict1 is dict2 ? False

但浅拷贝完，两个变量中的元素的值是一样的。

list1==list2 ? True
dict1 == dict2 ? True
set1==set2 ? True

2.2 对于列表

对于列表，还可以通过切片操作符“：”来完成浅拷贝

list1 = [1, 2, 3]
list2 = list1[:]
print(list2)
print("list1 == list2 ?",list1 == list2)
print("list1 is list2 ?",list1 is list2)

[1, 2, 3]
list1 == list2 ? True
list1 is list2 ? False

2.3 使用 copy.copy() 函数

函数 copy.copy() 函数，适用于任何数据类型

import copy

list1 = [1, 2, 3]
list2 = copy.copy(list1)
print(list2)
print("list1 == list2 ?",list1 == list2)
print("list1 is list2 ?",list1 is list2)

set1 = {1, 2, 3}
set2 = copy.copy(set1)
print(set2)
print("set1 == set2 ?",set1 == set2)
print("set1 is set2 ?",set1 is set2)

dict1 = {1:'xiaoming', 2:'xiahua',3:'xiaoli'}
dict2 = dict(dict1)
print(dict2)
print("dict1 == dict2 ?",dict1 == dict2)
print("dict1 is dict2 ?",dict1 is dict2)

[1, 2, 3]
list1 == list2 ? True
list1 is list2 ? False

{1, 2, 3}
set1 == set2 ? True
set1 is set2 ? False

{1: 'xiaoming', 2: 'xiahua', 3: 'xiaoli'}
dict1 == dict2 ? True
dict1 is dict2 ? False

2.4 对于元组

对于元组，使用 tuple() 或者切片操作符 ‘:’ 不会创建一份浅拷贝，相反它会返回一个指向相同元组的引用：

tuple1 = (1, 2, 3)
tuple2 = tuple(tuple1)
print(tuple2)
print("tuple1 == tuple2 ?",tuple1 == tuple2)
print("tuple1 is tuple2 ?",tuple1 is tuple2)

tuple1 = (1, 2, 3)
tuple2 = tuple1[:]
print(tuple2)
print("tuple1 == tuple2 ?",tuple1 == tuple2)
print("tuple1 is tuple2 ?",tuple1 is tuple2)

(1, 2, 3)
tuple1 == tuple2 ? True
tuple1 is tuple2 ? True

(1, 2, 3)
tuple1 == tuple2 ? True
tuple1 is tuple2 ? True

使用 tuple() 或者切片操作符 ‘:’ 不会创建一份浅拷贝，因为它开辟新的内存存储的是原对象的引用，而没有创建新的对象来存储原对象的子对象的引用，所以不是浅拷贝。相反它会返回一个指向相同元组的引用。

对字符串使用 str() 或者切片操作符 ‘:’，原理和 元组相同。

str1 = 'operation'
str2 = str1[:]
print(str2)
print("str1 == str2 ?",str1 == str2)
print("str1 is str2 ?",str1 is str2)

operation
str1 == str2 ? True
str1 is str2 ? True

str1 = 'operation'
str2 = str(str1)
print(str2)
print("str1 == str2 ?",str1 == str2)
print("str1 is str2 ?",str1 is str2)

operation
str1 == str2 ? True
str1 is str2 ? True

import copy

set1 = (1, 2, 3)
set2 = copy.copy(set1)
print(set2)
print("set1 == set2 ?",set1 == set2)
print("set1 is set2 ?",set1 is set2)

(1, 2, 3)
set1 == set2 ? True
set1 is set2 ? True

import copy

set1 = 'operation'
set2 = copy.copy(set1)
print(set2)
print("set1 == set2 ?",set1 == set2)
print("set1 is set2 ?",set1 is set2)

operation
set1 == set2 ? True
set1 is set2 ? True

也就是说，对字符串和元组使用 copy（）、[:]、本身的构造器完成的复制，都只是开辟了内存存储原对象的引用，而不是存储原对象的子对象的引用。

2.5 关于切片操作符 ‘:’

切片操作符 ‘:’ 不能用于字典和集合完成浅拷贝

# set1 = {1, 2, 3}
# set2 = set1[:]

# dict1 = {1:1, 2:2, 3:3}
# dict2 = dict1[:]

2.6 和赋值的区别

和赋值的本质区别在于，赋值只是把原对象的引用给到新对象

set1 = {1:1, 2:2, 3:3}
set2 = set1
print(set2)
print("set1 == set2 ?",set1 == set2)
print(id(set1))
print(id(set2))

{1: 1, 2: 2, 3: 3}
set1 == set2 ? True
2515252654368
2515252654368

2.7 浅拷贝需注意的问题

对数据采用浅拷贝的方式时，如果原对象中的元素不可变，那倒无所谓；但如果元素可变，浅拷贝通常会出现一些问题，

list1 = [[1, 2], (30, 40)]
list2 = list(list1)
list1.append(100)
print("list1:",list1)
print("list2:",list2)
list1[0].append(3)
print("list1:",list1)
print("list2:",list2)
list1[1] += (50, 60)
print("list1:",list1)
print("list2:",list2)

list1: [[1, 2], (30, 40), 100]
list2: [[1, 2], (30, 40)]
list1: [[1, 2, 3], (30, 40), 100]
list2: [[1, 2, 3], (30, 40)]
list1: [[1, 2, 3], (30, 40, 50, 60), 100]
list2: [[1, 2, 3], (30, 40)]

2、深拷贝

Python 中以 copy.deepcopy() 来实现对象的深度拷贝

import copy

list1 = [[1, 2], (30, 40)]
list2 = copy.deepcopy(list1)

list1.append(100)
print("list1:",list1)
print("list2:",list2)

list1[0].append(3)
print("list1:",list1)
print("list2:",list2)

list1[1] += (50, 60)
print("list1:",list1)
print("list2:",list2)

list1: [[1, 2], (30, 40), 100]
list2: [[1, 2], (30, 40)]
list1: [[1, 2, 3], (30, 40), 100]
list2: [[1, 2], (30, 40)]
list1: [[1, 2, 3], (30, 40, 50, 60), 100]
list2: [[1, 2], (30, 40)]